Creatine vs Peptides: What the Trial Count Shows

The comparison that sells peptides is against nothing. A vendor page describes what a compound does in rats and lets the reader supply the benchmark. Set an actual benchmark next to it — the cheapest, most-studied ergogenic aid in the building — and the shape of the argument changes, in both directions. Creatine's advantage is enormous in evidence and small in effect. This page reports both halves, as reporting on the literature — it tells no one what to use (our standards).

The count

Searched on PubMed, 2026-09-06.

The ratio between the first row and the rest is the argument. It is also the reason this site's our count of the human muscle-growth evidence scores creatine 5 and the injectables 1 or 2 — the rubric rewards replicated human trials, and only one column of this table has any.

Now the uncomfortable half: how big is creatine's effect?

Small. This deserves saying clearly, because the supplement industry overstates it and because a publication that only polices peptide claims is not being consistent.

The largest meta-analysis of the classic literature pooled 100 placebo-controlled studies and calculated effect sizes across 163 body-composition outcomes. The result: 0.17 for body composition, 0.24 for tasks lasting 30 seconds or less, 0.19 for glycolytic work. The paper's own conclusions note that the effect was larger for upper-body than lower-body exercise and that creatine "does not appear to be effective in improving running and swimming performance" (Branch, International Journal of Sport Nutrition and Exercise Metabolism 2003).

The modern estimate, restricted to trials that measured muscle directly with MRI, CT or ultrasound rather than inferring it from body mass, is smaller. Burke and colleagues analysed 44 outcomes across 10 studies of at least six weeks and found a pooled estimate of 0.11, with a 95% credible interval from -0.02 to 0.25 — "a very small effect favoring creatine supplementation when combined with resistance training" — and site-specific gains of 0.10 to 0.16 cm in muscle thickness (Burke et al., Nutrients 2023, PMC10180745).

A tenth of a centimetre of muscle thickness, with a credible interval that touches zero, is not a transformation. Tracer studies add that creatine does not make muscle build protein any faster; how it adds muscle through training instead is the subject of its own page. It is a real, replicated, cheap and legal edge, and it is the largest such edge anyone in this comparison can document.

Who paid for the creatine literature

Worth stating, because it is the objection a fair reader should raise. The 2023 imaging meta-analysis carries a conflict-of-interest statement disclosing a former scientific advisory board role at a supplement manufacturer that sells creatine, industry-sponsored creatine research, creatine donations and conference travel support, expert-witness and consultancy work in creatine litigation, and sales of creatine educational videos. The 2021 review of creatine in women's health discloses that two of its four authors advise a creatine manufacturer, with the note that the company had no role in the content.

These disclosures are printed in the papers and are easy to check. Two observations follow. The findings they produced are conspicuously unflattering — an effect size of 0.11 with an interval crossing zero is not what a sponsored analysis usually delivers. And the comparison compound has no equivalent disclosure to examine, because it has no trials to disclose interests in. An absent literature cannot have a conflict of interest, and that is not a point in its favour.

The same disclosure test applies to the safety side, where it cuts in the industry's favour and should be named anyway. The 685-trial analysis that finds no excess of side effects over placebo, and the 2025 trial that found no effect on hair follicles, both carry substantial creatine-industry conflict statements. Neither result is what a sponsor commissions, and both are checkable against the underlying trials — but the reader is entitled to the same footnote we would print about a vendor-funded peptide study.

What the peptide side would need to show

Not more mechanism. A randomised, placebo-controlled trial in trained adults, with a strength or muscle-size endpoint measured by imaging, powered to detect something. That trial has not been run for BPC-157, TB-500, CJC-1295, ipamorelin, sermorelin or IGF-1 LR3. It was run for anabolic steroids, in 1996, and our peptides-versus-steroids page is about what follows from that asymmetry.

The nearest approach is instructive. The only randomised human trial of ipamorelin asked whether it got the bowel moving again after surgery — nothing to do with muscle — and it missed: patients reached a first tolerated meal at a median 25.3 hours on the drug and 32.6 on placebo, a gap that was not significant (p = 0.15, 114 patients analysed) (Beck et al., International Journal of Colorectal Disease 2014). The class has one properly randomised, placebo-controlled human result and it is null.

Growth hormone itself, the thing all the secretagogues are marketed as a route to, has been tested in healthy adults and the outcome is the ceiling for the whole category: a meta-analysis of 18 study populations and 220 treated participants found fat mass down 2.1 kg and lean body mass up 2.1 kg, with weight unchanged, alongside significantly increased rates of soft-tissue oedema, arthralgia, carpal tunnel syndrome and gynecomastia. Its conclusion was that growth hormone "cannot be recommended" for this use (Liu et al., Annals of Internal Medicine 2007). No strength outcome was synthesised. MuscleLedger's page on muscle-and-fat-loss claims works through what that 2 kg each way does and does not mean.

The part that is not about efficacy at all

For anyone in a tested pool, the comparison ends before the effect sizes. Creatine is not on the WADA Prohibited List. Every peptide in this comparison is: BPC-157 under S0 (non-approved substances), TB-500 under S2.3 with thymosin-beta-4, and the secretagogues under S2.2, banned year-round rather than only on competition day — and the NCAA's 2026-27 banned classes catch the same molecules. Whether a given test would actually find them is a separate matter, set out on peptides and drug testing; the answer changes nothing about the prohibition.

The supply side differs too. Creatine monohydrate is a commodity food supplement. The research compounds reach buyers through a research-use-only channel that a 2026 review in Sports Medicine calls "a parallel 'gray market' of unapproved compounds… operating largely outside of regulatory oversight", with "rigorous human safety data… scarce" (Mendias and Awan, Sports Medicine 2026). Purity and identity are not verified by anyone in that chain by default.

The honest summary

Creatine's evidence advantage over every research compound in this category is close to total: thousands of randomised human trials against zero. The same gap opens on the aerobic side, where the endurance compounds trace back to a single 2008 experiment in mice that were not exercising. Creatine's effect advantage over doing nothing is small and getting smaller as measurement improves. Both of those are true at once, and a page that reports only the first is selling something.

That scepticism has to run all the way down, including into the parts of the creatine story that everyone repeats. The loading protocol on every label came from two biopsy studies in 48 men that measured no performance outcome at all, and the study that established loading also showed a slower intake reaching the same saturation. Women are missing from nearly half the trials. And the evidence thins to almost nothing as soon as the compound leaves the tub: the gummy format rests on one unblinded study, while lab testing of gummy brands found about half short of the creatine they printed. Applying a lower standard to the cheap legal option than to the expensive illegal one would be the same failure in the opposite direction. That standard, applied across every claim made for the supplement at once, produces a ranking rather than a verdict: creatine's endpoints graded from strong to barely studied, with the cognitive claim the one a European regulator declined to authorise in 2024. Every other claim this site has scored sits in the same format on the full list of ledger rows, numbered in the order each was researched.

Sources and dates

• PubMed counts run 2026-09-06 with the search terms named in the table.
• Branch, International Journal of Sport Nutrition and Exercise Metabolism 2003 — meta-analysis of 100 studies
• Burke et al., Nutrients 2023 — direct-imaging meta-analysis (PMC10180745)
• Smith-Ryan et al., Nutrients 2021 — creatine in women's health (PMC7998865)
• Beck et al., International Journal of Colorectal Disease 2014 — ipamorelin phase 2
• Liu et al., Annals of Internal Medicine 2007 — growth hormone in the healthy elderly
• Mendias and Awan, Sports Medicine 2026
• WADA 2026 Prohibited List, sections S0, S2.2 and S2.3; NCAA 2026-27 banned-drug classes.